This paper describes MAgentA (Musical Agent Architecture), which is an Agent that automatically composes background music in real time using the emotional state of the environment within which it is embedded. Its goal is to generate a "film-like music" output for a virtual environment emphasizing the dramatic aspects of such environment.
The approach followed in MAgentA is that of dynamically choosing an appropriate composition algorithm, among several others stored in a database, which were previously processed and duly associated to a particular emotional state.
MAgentA is in the process of being integrated from an abstract level in the FantasyA Virtual Environment, a wide scale virtual environment game currently being researched with INESC.
[1]
ChÅng Yu.
Computer generated music composition
,
1996
.
[2]
Judy Robertson,et al.
Real-time music generation for a virtual environment
,
1998
.
[3]
Curtis Roads,et al.
Research in music and artificial intelligence
,
1985,
CSUR.
[4]
Sho Yoshida,et al.
Automatic background music generation based on actors' mood and motions
,
1994,
Comput. Animat. Virtual Worlds.
[5]
A. North,et al.
Liking, arousal potential, and the emotions expressed by music.
,
1997,
Scandinavian journal of psychology.